obstacle to approve new materials for commercial MP
applications has to do with membrane resiliency regarding
real processing conditions and membrane capability to
maintain its characteristics (e.g., selectivity and capacity)
through reasonable operation times.
1.1 Software for Simulation of MP Units in Gas
Processing
Currently, as far as we can see, there are no commercial
computational tools available for rigorous design and simulation of general membrane-permeation units. When existent, such type of software is normally developed for local
and restricted ad hoc finalities of MP developers, MP manufacturers, and certain MP users. On the other hand,
oil-and-gas companies, which have to operate large-capacity
processing plants of CO 2 -rich NG at offshore sites, are
experiencing a crucial dependence on such category of
simulation and design tools.
These MP modeling tools are necessary, for example, to
revamp operating MP units in order to accommodate—in a
new processing flowsheet—new raw NG flow rates, new
(higher) CO 2 contents in raw NG, and new (stricter) CO 2
separation targets. Accurate MP models are also necessary
for daily supervision of operating MP plants, particularly,
regarding loss prevention and safety because membrane
cartridges can burst with certain frequency during the lifetime of MP units for high-pressure NG processing (Bernardo
et al. 2009).
However, since MP units have a large number of specific
configurational details—e.g., geometric aspects (diameters,
lengths, thicknesses), material bulk properties (density, heat
capacity, operational limits), external and internal heat
transfer coefficients (trans-membrane and trans-shell),
permeate/retentate flow/contact configurations (parallel
flows,
countercurrent
flows,
crossed
flows),
permeate/retentate locations relative to the membrane
(inside/outside), material surface properties (roughnesses,
specific areas), permeate/retentate head-loss parameters,
species permeances, etc.—the development of a truly rigorous steady-state MP simulator is a hard-core task, not
counting the thermodynamic aspects of non-isothermal,
non-isobaric, composition-changing permeate/retentate
compressible flows, and the geometrical/mathematical
issues characteristic of one-dimensional (1D) or
two-dimensional (2D) frameworks.
Moreover, even if such a MP simulator could be available, several materials/structural parameters of MP units for
high-pressure NG processing are not constant and expressively change with service time (Baker 2004). Some
changing parameters are evidently related to the degradations that the membrane material experiences through its
lifetime; namely, CO 2 /CH 4 selectivity, chemical stability,
structural resiliency, and mechanical stability. Such properties are known to change drastically with time in
high-pressure NG processing, always toward performance
deterioration, and eventually culminating with bursts due to
loss of structural stability or irreversible swelling and plasticization due to excessive intake of CO 2 and H 2 S into the
membrane body (Bernardo et al. 2009; Ebner and Ritter
2009).
Such difficulties to model membrane permeators are also
encountered with regard to other membrane-based unit
operations in the context of acid-gas removal from CO 2 -rich
NG. This is the case, for example, of gas–liquid membrane
contactors or GLMC (Marzouk et al. 2010), which are
membrane operations for CO 2 removal from NG where
permeation through skin-dense membranes is not the main
mechanism. Instead, the high-pressure retentate gas is contacted with an alkaline solvent at similar (slight lower)
pressure (e.g., aqueous-MEA or aqueous-MDEA) through a
porous membrane. In this case, the selectivity is imposed by
the aqueous alkaline solvent not by the membrane.
The GLMC also bears analogous critical modeling aspects
where flow geometry, flow/contact configurations, compressible multi-phase and single-phase flows, thermodynamic property calculations, transport phenomena, and
chemical multi-reaction equilibrium play important roles.
Recently, de Medeiros et al. (2013b) presented an
equilibrium-based rigorous model for simulation of gas–
liquid membrane contactors (GLMC) operating with
hollow-fiber porous membranes (HFM) for CO 2 removal
from high-pressure NG with aqueous equimolar solutions of
Table 1 Manufacturers of
cellulose-acetate membranes for
NG processing
Manufacturer
Membrane
type
Element
orientation
Element
L Â D
Element
installation
Gas contact
UOP
SWM*
Horizontal
1 m  0.2 m
Tandem elements
in tubes
Cross-flow
NATCO
Schlumberger
HFM*
Vertical
2 m  0.4 m
Several
Single element
Cross-flow
Parallel-flow
Air liquid
HFM
Vertical
Horizontal
Several
Single element
Cross-flow
Parallel-flow
*SWM Spiral-wound membrane; HFM Hollow-fiber membrane
146
J. L. de Medeiros et al.
applications has to do with membrane resiliency regarding
real processing conditions and membrane capability to
maintain its characteristics (e.g., selectivity and capacity)
through reasonable operation times.
1.1 Software for Simulation of MP Units in Gas
Processing
Currently, as far as we can see, there are no commercial
computational tools available for rigorous design and simulation of general membrane-permeation units. When existent, such type of software is normally developed for local
and restricted ad hoc finalities of MP developers, MP manufacturers, and certain MP users. On the other hand,
oil-and-gas companies, which have to operate large-capacity
processing plants of CO 2 -rich NG at offshore sites, are
experiencing a crucial dependence on such category of
simulation and design tools.
These MP modeling tools are necessary, for example, to
revamp operating MP units in order to accommodate—in a
new processing flowsheet—new raw NG flow rates, new
(higher) CO 2 contents in raw NG, and new (stricter) CO 2
separation targets. Accurate MP models are also necessary
for daily supervision of operating MP plants, particularly,
regarding loss prevention and safety because membrane
cartridges can burst with certain frequency during the lifetime of MP units for high-pressure NG processing (Bernardo
et al. 2009).
However, since MP units have a large number of specific
configurational details—e.g., geometric aspects (diameters,
lengths, thicknesses), material bulk properties (density, heat
capacity, operational limits), external and internal heat
transfer coefficients (trans-membrane and trans-shell),
permeate/retentate flow/contact configurations (parallel
flows,
countercurrent
flows,
crossed
flows),
permeate/retentate locations relative to the membrane
(inside/outside), material surface properties (roughnesses,
specific areas), permeate/retentate head-loss parameters,
species permeances, etc.—the development of a truly rigorous steady-state MP simulator is a hard-core task, not
counting the thermodynamic aspects of non-isothermal,
non-isobaric, composition-changing permeate/retentate
compressible flows, and the geometrical/mathematical
issues characteristic of one-dimensional (1D) or
two-dimensional (2D) frameworks.
Moreover, even if such a MP simulator could be available, several materials/structural parameters of MP units for
high-pressure NG processing are not constant and expressively change with service time (Baker 2004). Some
changing parameters are evidently related to the degradations that the membrane material experiences through its
lifetime; namely, CO 2 /CH 4 selectivity, chemical stability,
structural resiliency, and mechanical stability. Such properties are known to change drastically with time in
high-pressure NG processing, always toward performance
deterioration, and eventually culminating with bursts due to
loss of structural stability or irreversible swelling and plasticization due to excessive intake of CO 2 and H 2 S into the
membrane body (Bernardo et al. 2009; Ebner and Ritter
2009).
Such difficulties to model membrane permeators are also
encountered with regard to other membrane-based unit
operations in the context of acid-gas removal from CO 2 -rich
NG. This is the case, for example, of gas–liquid membrane
contactors or GLMC (Marzouk et al. 2010), which are
membrane operations for CO 2 removal from NG where
permeation through skin-dense membranes is not the main
mechanism. Instead, the high-pressure retentate gas is contacted with an alkaline solvent at similar (slight lower)
pressure (e.g., aqueous-MEA or aqueous-MDEA) through a
porous membrane. In this case, the selectivity is imposed by
the aqueous alkaline solvent not by the membrane.
The GLMC also bears analogous critical modeling aspects
where flow geometry, flow/contact configurations, compressible multi-phase and single-phase flows, thermodynamic property calculations, transport phenomena, and
chemical multi-reaction equilibrium play important roles.
Recently, de Medeiros et al. (2013b) presented an
equilibrium-based rigorous model for simulation of gas–
liquid membrane contactors (GLMC) operating with
hollow-fiber porous membranes (HFM) for CO 2 removal
from high-pressure NG with aqueous equimolar solutions of
Table 1 Manufacturers of
cellulose-acetate membranes for
NG processing
Manufacturer
Membrane
type
Element
orientation
Element
L Â D
Element
installation
Gas contact
UOP
SWM*
Horizontal
1 m  0.2 m
Tandem elements
in tubes
Cross-flow
NATCO
Schlumberger
HFM*
Vertical
2 m  0.4 m
Several
Single element
Cross-flow
Parallel-flow
Air liquid
HFM
Vertical
Horizontal
Several
Single element
Cross-flow
Parallel-flow
*SWM Spiral-wound membrane; HFM Hollow-fiber membrane
146
J. L. de Medeiros et al.